by Mark Pagel
But modern genomic studies have begun to supply yet another answer to why we might allow at least some of the junk to build up in our genomes. It appears our genes use it to help them make our bodies. Our meager supply of genes is not capable on its own of specifying all the necessary parts to build a body, much less the precise times they are needed and all of the connections these parts make with each other. Genes make proteins, and proteins are the building blocks of our bodies. Our hair, muscles, nerves, fingernails, blood cells, and skin are all made of different kinds of proteins. Nearly every cell in your body carries a complete copy of your genome, but only some of your genes are used in any given place in your body and at any given time. This means there must be something that tells the genes that make your eyes not to switch on in the back of your head, or genes for teeth to stay silent in your toes. Something has to provide the instructions to get genes to team up to produce complex structures such as hearts and kidneys, or the chemical networks that create our metabolism. There is no little homunculus perched on a stool inside us calling out instructions from a manual. Instead, the vast quantities of junk DNA in our genomes fortuitously seem to contain a nearly unlimited variety of different messages whose language our genes have learned.
The complexity of our bodies might be built, then, on the enormous expansion of junk DNA that began hundreds of millions of years ago in our distant ancestors. Careful research shows that by complicated chemical means a gene can use stretches of the junk DNA to regulate or control when, where, and how much it is expressed. A gene is expressed when it makes a protein, and just as a person uses language to express a thought, genes seem to use the vast vocabulary of sequences of junk DNA to express themselves in exactly the right amounts and at exactly the right times and places in our bodies. Our genes are using the junk DNA to promote their interests within our bodies because genes that make better bodies are more likely to survive and be passed on to the next generation. Incredible as it might sound, junk DNA has been a source of great evolutionary innovation, and we all might just owe our existence, at least in part, to it. Differences in gene regulation are why two animals like chimpanzees and humans can be about 98.5 percent identical in the sequences of their genes and yet be so different on the outside. Junk DNA might even have enabled the ancient biological transition from single-celled organisms like yeast to complex multicellular organisms such as ourselves, elephants, lizards, and monkeys. Yeast and other single-celled organisms don’t have arms and legs, brains and eyes, and it turns out they have very little junk DNA. This is not to say junk DNA exists for our good, or even for the gene’s good. It exists because it is good at making copies of itself, and genes have merely been able to exploit its presence.
Before we can grasp the significance of gene regulation to the story of human language, we need to understand one more feature of our genetic system. That feature is that our DNA is a digital system of inheritance. A digital signal is one, like Morse code, in which information is transmitted in distinct packets. Morse code relies on just two of these packets, a dot and a dash, and for that reason is a binary digital signal. Our DNA’s digital signal, rather than having just two packets, uses four distinct molecules called bases or nucleotides, and these are normally abbreviated A, C, G, and T. Digital signals are needed whenever a system requires great fidelity and great variety. The first of these, fidelity, refers to the ability to be transmitted over and over without losing the signal. Digital signals have fidelity because they have a measure of being self-correcting. If when transmitting a dot in Morse code a little bit of error creeps in, the receiver can usually recognize from the context of the broader message that the signal is still a dot. This means that when that dot is transmitted again, the error will have been removed.
By comparison, if the signal to be transmitted is a continuous rather than digital one, receivers cannot know what part is error and what part is the true signal. Loudness, brightness, or length are all continuous signals, and if, for example, I send a loud signal to you, and make it just a bit too loud, you will not know that I have made an error and thus you will be likely to repeat the error if you transmit the signal to someone else. Over long periods of time and many transmissions of this signal, it can drift well away from its starting point because no receiver ever knows what part of the signal is correct and what part is error. Similarly, if I ask you to copy a drawing I have made, you might be able to produce something like it, but it will not be an exact copy. But if I made it from a set of different-colored dots—like the Georges Seurat pointillist paintings—you probably could copy it exactly, even if with some difficulty if it were a Seurat.
There is a beauty, then, in the simplicity of our genetic system. Its digital nature means we can expect it to have high fidelity, and this is just what we require of a system that has to be transmitted repeatedly over millions and millions of years. But we also require our genetic systems to have variety or the ability to produce a wide range of different messages. And here digital systems excel. A Morse code message of just two of its binary signals already yields four possible combinations, each one of which is a different message (dot-dot, dot-dash, dash-dot, and dash-dash). A message of length three can specify eight different instructions (2 × 2 × 2); there are sixteen possibilities with just four signals and thirty-two with five, and in general a message of length n can produce 2n (two multiplied by itself n times) varieties. A message of just twenty binary signals can therefore specify over a million different outcomes (1,048,576, to be exact) and each message is easy to distinguish from the others.
A DNA signal of length n can produce 4 raised to the power n different messages. This is because every position in the gene can now take four different outcomes (A, C, G, or T). By stringing together long chains of these bases, our genes can make a variety of different proteins. Equally, though, this means the vast stores of junk DNA in our genomes can also produce an effectively unlimited variety of different instructions for regulating our genes, and this might be just what our bodies need to function effectively. In fact, a measure of the importance of the digital regulatory sequences that reside in our junk DNA is that their precise sequences are often highly conserved. Thus, for example, a mouse and an elephant might share some exact or nearly exact regulatory sequences, despite being separated by many millions of years of evolution.
Now we are in a position to understand what is so special about human language. Like our genes and Morse code, human language is also a digital communication system. No other animal’s is, and if I have done my work in preparing you, you will realize this tells us that human language, unlike the continuously varying signals of the other animals, must have evolved for a task that requires both great fidelity and great variety. Our language is digital by virtue of being built from discrete entities we call words, and those words are made from simpler building blocks of discrete sounds—the DNA of our language—called phonemes. The digital information that forms a word is sent through the air in pressure waves that we form with our tongue, mouth, and breath. Your ear upon hearing one of them sends the signal to your brain, where it is decoded, just as your television decodes the signals it receives from your wireless remote control, or your body decodes the message in a gene. The system to a large extent “snaps back” into place or is self-correcting when small errors are made because if a word is mispronounced or given an unusual accent or emphasis, receivers can still normally work out what word was intended, and thus this alteration is not retained in the next transmission of that word.
Being digital, our language can easily transmit millions upon millions of different messages, not—like most animal signals—limited to just a few dimensions such as bigger or smaller, louder or softer, or more or less fierce. Imagine you utter just a three-word sentence with a subject, a verb, and an object—such as I kicked [the] ball. If you can choose from 20 different words at each position of that sentence, you already can make 8,000 different sentences (20 × 20 × 20). Of course, there are many mo
re words than this, and sentences can be of any length. This means our linguistic systems can effortlessly produce many more messages than the digital color systems of your television or computer can produce different colors on their screens. In fact, our language can produce an unlimited variety of messages because a sentence can be of any length.
Edmund Rostand’s engaging play Cyrano de Bergerac, about a seventeenth-century French nobleman and duelist with a large nose, provides an amusing illustration of the nuanced expression language can achieve—just the right expression for every circumstance. When the Viscount Valvert tells Cyrano, “Your nose is a trifle large,” Cyrano, amused but indignant, responds:
Young sir you are too simple, you might have said at least a hundred things by varying the tone… like this, suppose… . Aggressive: “Sir, if I had such a nose I’d amputate it!” Friendly: “When you sup it must annoy you, dipping in your cup; You need a drinking-bowl of special shape!” Descriptive: “’Tis a rock! . . . a peak! . . . a cape!—A cape, forsooth! ’Tis a peninsular!” Curious: “How serves that oblong capsular? For scissor-sheath? Or pot to hold your ink?” Gracious: “You love the little birds, I think. I see you’ve managed with a fond research to find their tiny claws a roomy perch!” Truculent: “When you smoke your pipe… suppose that the tobacco-smoke spouts from your nose—Do not the neighbors, as the fumes rise higher, cry terror-struck: ‘The chimney is afire’?” Considerate: “Take care… your head bowed low by such a weight… lest head o’er heels you go!” Tender: “Pray get a small umbrella made, lest its bright color in the sun should fade!” Pedantic: “That beast Aristophanes names Hippocamelelephantoles must have possessed just such a solid lump of flesh and bone, beneath his forehead’s bump!” Cavalier: “The last fashion, friend, that hook? To hang your hat on? ’Tis a useful crook!” Emphatic: “No wind, O majestic nose, can give THEE cold!—save when the mistral blows!” Dramatic: “When it bleeds, what a Red Sea!” Admiring: “Sign for a perfumery!” Lyric: “Is this a conch? . . . a Triton you?” Simple: “When is the monument on view?” Rustic: “That thing a nose? Marry-come-up! ’Tis a dwarf pumpkin, or a prize turnip!” Military: “Point against cavalry!” Practical: “Put it in a lottery! Assuredly ’twould be the biggest prize!” Or… parodying Pyramus’ sighs… “Behold the nose that mars the harmony of its master’s phiz! blushing its treachery!”—Such, my dear sir, is what you might have said, had you of wit or letters the least jot… .
“Wit or letters the least jot” is precisely what Cyrano himself would have lacked had he been any other animal apart from a human because none of them is blessed with our digital communication system. Jared Diamond describes how in America in the late 1940s, Keith and Catherine Hayes brought up a chimpanzee called Viki in their household. But instead of bringing her up as a chimpanzee, they attempted to raise her like a human infant. They played and talked to her, fed her at the table, sang to her, and generally just treated her like another member of the family. Viki excelled at playtime but struggled despite heroic efforts on the part of her owners ever to speak. Even after hours and hours of intensive training she could manage just three words, “mama,” “papa,” and “cup,” and then it must be acknowledged that the Hayeses were better than outsiders at recognizing these “words”—something to which any proud parent of a two-year-old will also have to confess.
Many years later, Diamond imagined Viki trying to speak, using at most the two vowels and two consonant sounds in these words (charitably, Viki could produce four phonemes, an a and u, and a c and p, and even this probably overestimates a normal chimpanzee’s abilities). Diamond, who was giving a lecture at Trinity College in Dublin, asked his audience to “try seeing how many different words you could speak if you could only pronounce the vowels a and u, and the consonants c and p. If you wanted to say ‘Trinity College is a fine place to work,’ all you could manage would be ‘Capupa Cappap up a cap capcupap.’ Your attempt to say ‘Trinity College is a bad place to sneeze’ would result in identical sounds.”
What task requires the fidelity and variety that human language possesses? Our social phenotypes or cultures are immensely complex compared to those of any other animal, and we derive gains from varying our “expression” inside it much like a gene does inside a body. Thus, we deploy our digital language to manipulate this social phenotype in ways that promote our survival and reproduction, and in the same way that genes use their vast digital repository of junk DNA to vary their expression inside our physical bodies. In fact, we could say that deep down language might just be the latest form of gene regulation—the voice of our genes. We rely on the great variety and vocabulary of language to share ideas, to promote cooperation, to build alliances, and to enhance our reputations, just as genes rely on the great vocabulary and variety of junk DNA to regulate how much, when, and where in our bodies they are expressed. We use our language to be deceptive or charming, kind and forgiving, or spiteful and vindictive. We use it to manipulate or bewitch others, to collude with them, or to foster or defuse factional disputes. We use language to embroider and exaggerate our own dossiers and gently diminish or disparage those of others. We alter our speech or dialect strategically—linguists call this code switching—to signal our connection to a group or individual.
As we saw with Viki, none of this could be achieved with the simple continuously varying signals of the rest of the animal kingdom, limited as they are to “more” or “less” of some quantity. Instead we needed a digital mechanism capable of great variety and fidelity. And so, just as junk DNA might have enabled the transition to large complex multicellular organisms, the evolution of language might have enabled the transition to our complex societies. Not being able to speak in the newly emerging human societies would have been like being a bird that could not fly. Just as wings open up an entirely new sphere to be exploited, language opened up the sphere of cooperation, and genes for human language would have quickly spread.
Before leaving the topic of digital systems and regulation, I would like to mention that it was yet another kind of digital regulation that enabled our modern electronics revolution, and that has changed the way we lead our lives. John Mattick points out that up until recently airplanes, clocks, and even computers were analogue as opposed to digital devices, manipulated by continuously varying signals such as levers, springs, heat, or pressure. For example, airplanes were flown with a stick, and large springs were wound up to drive clocks. But once our engineers discovered digital regulation—instructions encoded in strings of binary numbers arbitrarily long, and hence precise—these machines and a host of new ones, such as digital cameras and music players, could become more complex.
The U.S. Air Force’s Stealth fighter planes provide an extreme example of a machine that could not exist without digital regulation. These planes achieve their stealth in part from having a shape that makes them reflect enemy radar in such a way that the radar image loses its coherence. The trouble is that this shape makes them virtually impossible to fly, and they rely on millions of split-second adjustments to keep them airborne. Making these adjustments is beyond the capabilities of mere (analogue) human pilots—so much so that the Stealth planes are flown only with a large input from on-board computers. Little did the designers of these machines know that nature had beaten them to this digital regulation by hundreds of millions of years.
WORDS, LANGUAGES, ADAPTATION, AND SOCIAL IDENTITY
IF LANGUAGE is “the voice of our genes,” we should see evidence of this important role in how we use language and in how elements of language evolve. One way to do this is to ask what words we use most often in our everyday speech, and when we do this, two surprising results emerge. One is that there is a huge disparity between how often different words get used, with some being used hundreds and even thousands of times more often than others. So great is this disparity that somewhere around 25 percent of all our speech is made up from a mere twenty-five words! According to the Oxford English Dictionary, the English language’s top
twenty-five include the, I, you, he, this, that, have, to be, for, and and; but they, we, say, and she make it into the top thirty. All of these are used thousands to perhaps 30,000 times per every million utterances, whereas most of the remaining 200,000 or more other English words that great dictionary catalogues tend to be used only very infrequently, some of them only a few times in every million utterances. When, for example, was the last time you used the words indefatigable or expository or behoove?
The second surprise is that speakers of different languages all use more or less this same subset of words frequently in their speech and a different set infrequently. It seems that around the world we all talk about the same things and in roughly the same amounts. This holds whether the speaker is French or Bantu, Chinese or Hungarian, Basque, English, Turkish, Finnish, Greek, or using a Polynesian language. We know this from studying a common set of words that the American linguist Morris Swadesh, working in the 1950s, proposed as a “fundamental vocabulary.” Swadesh’s goal was to identify a list of words that should be found in all human languages. It includes words such as what, where, when, mother, father, fish, bird, hold, count, throw, float, say, day, night, bite, eat, sky, drink, and louse, along with number words, names of body parts, pronouns, colors, and common verbs, adjectives, and nouns. It deliberately excludes technological words or words that describe specific environments or features of them. Louse might surprise modern readers, but Swadesh was interested in the history of language use, and lice infections have been a common feature of our history—and still are in many parts of the world (and annoyingly among children of school age).
